文献类型：会议论文

英文题名：Adaptive Remote Sensing Image Attribute Learning for Active Object Detection

作者：Xu, Nuo[1];Huo, Chunlei[1];Guo, Jiacheng[2];Liu, Yiwei[3];Wang, Jian[4];Pan, Chunhong[1]

第一作者：Xu, Nuo

通讯作者：Xu, N[1]

机构：[1]Chinese Acad Sci, Univ Chinese Acad Sci, Sch Artificial Intelligence, NLPR,Inst Automat, Beijing, Peoples R China;[2]Beijing Informat Sci & Technol Univ, Beijing, Peoples R China;[3]Beijing Univ Civil Engn & Architecture, Beijing, Peoples R China;[4]Beijing Union Univ, Coll Robot, Beijing, Peoples R China

第一机构：Chinese Acad Sci, Univ Chinese Acad Sci, Sch Artificial Intelligence, NLPR,Inst Automat, Beijing, Peoples R China

通讯机构：[1]corresponding author), Chinese Acad Sci, Univ Chinese Acad Sci, Sch Artificial Intelligence, NLPR,Inst Automat, Beijing, Peoples R China.

会议论文集：25th International Conference on Pattern Recognition (ICPR)

会议日期：JAN 10-15, 2021

会议地点：ELECTR NETWORK

语种：英文

摘要：In recent years, deep learning methods bring incredible progress to the field of object detection. However, in the field of remote sensing image processing, existing methods neglect the relationship between imaging configuration and detection performance, and do not take into account the importance of detection performance feedback for improving image quality. Therefore, detection performance is limited by the passive nature of the conventional object detection framework. In order to solve the above limitations, this paper takes adaptive brightness adjustment and scale adjustment as examples, and proposes an active object detection method based on deep reinforcement learning. The goal of adaptive image attribute learning is to maximize the detection performance. With the help of active object detection and image attribute adjustment strategies, low-quality images can be converted into high-quality images, and the overall performance is improved without retraining the detector.